Development of a program for diagnosing diseases by the test recognition method

Abstract

Theory of pattern recognition is an important theoretical and applied area in computer science. Depending on the nature of the task for recognition are used different approaches. In this article the authors investigate the current state of the use of supercomputers in the field of medical diagnostics and propose an affordable applied solution for diagnosing diseases based on the use of test recognition. Test recognition is based on such a combinatorial-logical approach as a test algorithm. Test algorithms are based on the analysis of a set of deadlock table tests and depend on the dimension of a given matrix of objects and their features. A deadlock test is an incompressible set of features that contains all the information about dividing a table into classes. In this work briefly presents the chronology of the development of test algorithms for solving recognition problems. The authors have developed an interpretation of the test algorithm to automate the diagnosis of diseases by identifying the proximity of the desired subset (available symptoms) to one of the existing ones (diagnoses) by defining deadlock tests. In addition to the presented algorithm, screenshots are presented with an example of a step-by-step solution in the form of a console application created in the high-level programming language C#. On the basis of the developed algorithm, the authors propose an example of a ready-made software solution in the form of a window application for diagnosing diseases based on the patient's symptoms. The developed software solution allows to mark the patient's symptoms and automatically calculate the most probable disease. The program narrows the range of acceptable values (diagnoses) depending on the symptoms marked by the user and provides an assessment of possible diseases, helping the doctor in making a diagnosis and minimizing errors made by the human factor. The article provides an assessment of the advantages and disadvantages of the developed solution, considers other areas of application of test recognition